Discrete Choice Analysis: Models, Estimation and Applications
An introduction to the main techniques of discrete choice analysis and the design of stated choice experiments.
20 June to 24 June 2016
Group discounts are available for three or more persons from the same organisation.
Please email: email@example.com for information about this.
An early bird rate is available until 2 May 2016.
Prices include GST
- Industry and Government: $3850.00
Early bird: $3272.50
- Academic: $2750.00
Early bird $2337.50
- Student: $2200.00
Early bird $1870.00
Almost without exception, everything human beings undertake involves a choice. In recent years there has been a growing interest in the development and application of quantitative statistical methods to study choices made by individuals or groups with the purpose of gaining a better understanding both of how choices are made and of forecasting future choice responses.
Discrete choice analysis and stated choice methods are widely used across diverse fields to study the behavioural responses of individuals, households and other organisations. This course is designed to provide both theory and practical experience in the building and estimating of simple (e.g., Multinomial Logit (MNL)) and more advanced choice models (e.g., mixed and generalised MNL, random parameter latent class), as well as in generating stated choice experimental designs. We will also cover future developments in the field of discrete choice analysis (e.g., nonlinear in parameters models, risk attitude, perceptual conditioning). Whilst theory will be covered, significant time will be spent in a computer lab, working on building models using real data, and generating simple surveys. Those completing the course will be capable of transferring the techniques taught to their own research areas.
This practical course will be useful for research across a broad range of fields in which consumer demand and choice is of interest, including: accounting, economics, engineering, environmental science, finance, health services, logistics, marketing, planning, transportation, and tourism. The course is intended for academics and practitioners in government and industry. An appreciation of basic statistical concepts is useful, but not essential for this course (please contact Andrew Collins if you have any concerns). Many previous attendees have come with no background in discrete choice modelling, but have completed the course at a level that has enabled them to immediately develop and estimate a range of choice models.
This is a practical and theory based course. We will be teaching how to estimate discrete choice models using software, and how to interpret the outputs using a real life data set for this. We will cover MNL, nested logit, latent class, error components, and (generalised) mixed logit (random parameters) for ordered and unordered choices for ranks, rates and choice responses. New developments in estimation such as estimation in willingness to pay space will also be discussed in the course. Widely used statistical tests, such as Delta, Wald and the Krinksy and Robb procedure, will also be covered. Furthermore, an introduction into the design of stated choice surveys will include the generation of orthogonal and efficient experimental designs. A variety of applications will be used to illustrate the techniques. The course will include presentations of the background theory for discrete choice modelling, different methods for combining survey data, and the most recently developed modelling techniques including nonlinear in parameters models, cumulative prospect theory models and process heuristics.
The focus of the course will be on the entire process, experimental design, model building, and model estimation (including of data definition, stated preference and revealed preference data). Recent advances in tools and methods have been used to model individual behaviour and to analyse market shares and change in demand in response to pricing and income and changes in available choice sets and choice characteristics.
Hands on problems with actual data sets will be used to augment the presentations. Applications in model estimation will be developed using NLOGIT 5.0 / LIMDEP 10.0 software (Econometric Software, Inc.). A 10% discount on the price for this software will be available to course participants wishing to purchase. Experimental designs will be generated using the Ngene 1.2 software (ChoiceMetrics Pty Ltd).
This course will be presented by four of the world’s leading academics in the field of discrete choice analysis:
Professor William (Bill) Greene
Bill is the Toyota Motor Corp Professor of Economics at the Stern School of Business, New York University. His fields of interest are applied econometrics, panel data analysis, discrete choice modelling, production economics, health econometrics, transport economics and planning, and economics of the entertainment industry. He is President of Econometric Software, Inc. and author of software LIMDEP and NLOGIT, textbooks Econometric Analysis (editions 1 to 7), books on discrete choice modelling, modelling ordered choices and applied choice analysis, and over 100 articles in peer reviewed journals.
Professor David Hensher
David is the Founding Director of the Institute of Transport and Logistics Studies at the University of Sydney. He is a fellow of the Academy of Social Sciences in Australia and recipient of a Lifetime Achievement Award from the International Association of Travel Behaviour Research, recognising his long-standing and exceptional contribution to IATBR and the wider travel behaviour community. David has published over 550 papers in key journals and books in transportation, economics and environmental science, and has been an active contributor to the choice modelling community for over 40 years; he was a pioneer in the introduction of stated choice methods, interactive agency choice experiments, and process heuristics such as attribute processing.
Professor Michiel Bliemer
Michiel is Professor of Transport and Logistics Network Modelling at the Institute of Transport and Logistics Studies at the University of Sydney. He has published over 200 articles in peer reviewed journals and conference proceedings, mainly on methodology in the areas of stated choice experimental design, travel behaviour, network modelling and traffic simulation. Michiel is co-developer of the Ngene software, which is the world leading software for generating designs for stated choice surveys, and in the past 10 years has taught many courses on experimental design in Europe and Australia.
Andrew is a Lecturer in Transport and Logistics Management at the Institute of Transport and Logistics Studies at the University of Sydney. He has broad research interests, spanning advanced discrete choice modelling methodology and its application across many fields, choice heuristics, stated choice experimental design, last mile logistics, freight transport, and land and air travel behaviour. Andrew's PhD, which examined techniques for handling attribute nonattendance in discrete choice models, was awarded the prestigious 2012 Eric Pas Dissertation Prize by the International Association for Travel Behaviour Research. Andrew is a co-developer of the Ngene software.
- Excellent course, excellent standard of delivery. Hard to imagine a better quality of course.
- Truly fantastic - very rewarding 5 days. Looking forward to applying all this new found knowledge to my research.
- This course was informative, useful and well-run. The material seems to be a good mixture of theory and practical and a good mixture of basic and more advanced topics.
- Very good indeed – I found the course extremely useful and entirely appropriate to my needs. Thank you.
- Fantastic course. Very well run and presented. I have learnt and been impressed. Thank you.
- Fantastic course for anyone interested in/practicing in discrete choice analysis. Course was very well structured and had a good balance between theory/lectures and hands-on workshops. Having the material presented by leaders in the field also meant that highly technical and difficult material was conveyed clearly and in an intuitive and easy (relatively!) to follow way.
- Great! Simply great!
- David, A wee note of thanks to you and your team for making my journey into discrete choice analysis an enjoyable one. You have a really good process, the presenters know their stuff, this is really state of the art in postgraduate/executive education. I found the content, scope and pace demanding, (which you want, the course would be a waste of time otherwise) but the structure and approach of the presenters nicely provided the remedy for this. Again, many thanks for a challenging and enjoyable week.
Dr Chris Batstone, Senior Resource and Environmental Economist, The Cawthron Institute, New Zealand
- I wanted to say thank you for running last week's choice modelling course. I'm really pleased I attended and am confident in taking my analysis forward more robustly. I've already started using the reference material and NLOGIT for some of my analysis… Thank you again for such a well-run, well-taught course; I'd recommend it to anyone doing choice modelling.
Anna Robak, PhD Candidate, Centre for Regulation and Market Analysis, University of South Australia
- Just thought I would drop you all a quick e-mail to say I thoroughly enjoyed the 5 day course. It's not very often I attend a course and I learn something in every session. I certainly did that with this course. Now all I have to do is apply what I have learnt by developing choice experiments for my research. Thanks to everyone for making us welcome and for looking after us so well over the five days.
Julia Logan, Head of Department, Patient Information Management Services, Child and Adolescent Health Service, Princess Margaret Hospital for Children, Perth